2033329616

If model has batch normalization layers, where should I use the bn_update()? ``` for _ in range(100): opt.zero_grad() loss_fn(model(input), target).backward() opt.step() opt.swap_swa_sgd() opt.bn_update(train_loader, model) ``` Is this setting correct？ Or...

In the file Lib/roi_pool.py, the boxes don't multiply spatial_scale, but the boxes is within the scope of the input image, but we need to get the feature of the roi...

In /lib/faster_rcnn_config.py, ``` # Classes: The types of objects the algorithm is trying to find # First class must be __background__ __C.CLASSES = ['__background__'] __C.NUM_CLASSES = 1 ``` Why is...

Can not find add button about header in auto label settings

2

How to reproduce the behaviour --------- Your Environment --------- * Operating System: win10 * Python Version Used: python3.9 * When you install doccano: 1.8.4 * How did you install doccano...

In dynamic_assign.py function dynamic_k_assign, there is a problem in assign zero when one anchor has more than one gt ``` def dynamic_k_assign(cost, pair_wise_ious): matching_matrix = torch.zeros_like(cost) ious_matrix = pair_wise_ious ious_matrix[ious_matrix...

I found a segmentation branch in the code, but in the paper there is no segmentation, and whether repeated experiments can achieve the same effect in paper if I remove...

``` -- Build files have been written to: /notebooks/caffe-master-new [ 1%] Built target proto [ 1%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile_1.dir/layers/cuda_compile_1_generated_DenseBlock_layer.cu.o /notebooks/caffe-master-new/src/caffe/layers/DenseBlock_layer.cu(271): error: too few arguments in function call detected...